Learning Photographic Global Tonal Adjustment with a Database of Input / Output Image Pairs

Vladimir Bychkovsky
Sylvain Paris
Adobe System, Inc
Eric Chan
Adobe System, Inc
Frédo Durand


Adjusting photographs to obtain compelling renditions requires skill and time. Even contrast and brightness adjustments are challenging because they require taking into account the image content. Photographers are also known for having different retouching preferences. As the result of this complexity, rule-based, one-size-fits-all automatic techniques often fail. This problem can greatly benefit from supervised machine learning but the lack of training data has impeded work in this area. Our first contribution is the creation of a high-quality reference dataset. We collected 5,000 photos, manually annotated them, and hired 5 trained photographers to retouch each picture. The result is a collection of 5 sets of 5,000 example input-output pairs that enable supervised learning. We first use this dataset to predict a user's adjustment from a large training set. We then show that our dataset and features enable the accurate adjustment personalization using a carefully chosen set of training photos. Finally, we introduce difference learning: this method models and predicts difference between users. It frees the user from using predetermined photos for training. We show that difference learning enables accurate prediction using only a handful of examples.



We are grateful to Katrin Eismann and Jeff Schewe for providing invaluable advice and for introducing us to the community of professional photographers. We thank Todd Carroll, David Mager, Jaime Permuth, LaNola Katheleen Stone, and Damian Wampler for their incredible patience while retouching thousands of photos. Special thanks to everyone who contributed their photos to this dataset: without you this work would not have been possible. This work was supported in part by grants from Foxconn and NSF (0964004) and a gift from Adobe.